The self-incompatibility system in Caltha palustris (Ranunculaceae)

Abstract

The investigated material consists of three plants, collected from two localities in the neighbourhood of Lund, four F1 families, totalling 42 plants (range 5–21 plants), and three backcross families, totalling 92 plants (range 19–40 plants). All plants were clearly self-incompatible, judging from seed-setting data in situ under greenhouse conditions. Among F, siblings (F, × Fj) there was no safe case of cross-incompatibilty (337 combinations tried out of the 592 possible). Parental plants pollinated by backcross families (P × BC) and reciprocally (BC × P) had, each, only one clearly documented case of cross-incompatibility (76 and 85 combinations tried, respectively, out of the 92 possible). Calculations were applied to these data, aiming at minimum degrees of cross-incompatibility to be expected in different intercrosses, at different numbers of S-loci, with disomic and tetrasomic inheritance, the assumptions being maximal S-allele diversity for the parental plants, independent inheritance, and chromosome assortment being random. Irrespective of the presence of disomic or tetrasomic inheritance, a team of complementary 5-genes had to be assumed. The increased range of seed set of parental plants when they were pollinated by backcross plants favoured an interpretation implying tetrasomic inheritance with low degrees of chromatid assortment for S-loci, substantially reduced frequencies of compatible pollen being expected to effect reduced seed set. Most likely, a gametophytic complementary 5-gene system, similar to that in the closely related genus Ranunculus, is successfully operating in the tetrasomic state in Caltha palustris.